TuRTLe Module Completion (NotSoTiny): leaderboard
TuRTLe module-completion leaderboard on the NotSoTiny benchmark, using formal equivalence and partition coverage over Tiny Tapeout-derived RTL modules.
Metric: Functionality EQV (imported). Source: huggingface.co. Source: benchmarklist.com. Saturation forecast: Around May 2027. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Kimi K2.5 | 31.57 | #144 |
| 2 | Gemma 4 31B (IT) | 29.54 | #176 |
| 3 | Qwen 3 Coder 480B A35B Instruct | 23.43 | #315 |
| 4 | GPT-OSS-120B | 20.9 | #341 |
| 5 | DeepSeek R1 0528 | 20.73 | #222 |
| 6 | Kimi K2 0905 | 19 | #288 |
| 7 | Qwen 2.5 72B Instruct | 14.7 | #454 |
| 8 | Qwen2.5-14B-Instruct-1M | 12.49 | #578 |
| 9 | Qwen 2.5 Coder 32B Instruct | 11.85 | #528 |
| 10 | Qwen 2.5 14B Instruct | 8.97 | #647 |
| 11 | Qwen 2.5 7B Instruct | 3.48 | #862 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=turtle-module-completion-notsotiny · How It Works · Data refreshed daily, snapshot 2026-10-20.